‘Deception Sleuth’ #Nature10 Scientist https://xitter.cf/t.co/xXL3Dnb4ZI • CS Prof. @UT3PaulSabatier @InstUnivFr • Scientific Text Mining • https://xitter.cf/t.co/7r0v4aXR0o
Toulouse, France
Joined January 2011
- Tweets17.8K
- Following1.5K
- Followers5K
- Likes31K
Pinned Tweet
The ‘Problematic Paper Screener’: flags unreliable passages in publications, supports a decentralised re-assessment of problematic papers, helps to decontaminate the scientific literature. irit.fr/~Guillaume.Cabanac/p… That's my research as @InstUnivFr member. leblob.fr/videos/guillaume-c…
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
los pollos hermanos 🐔
made with mitte.ai/app/hotel-lobby-cha…
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
Rather scary. On the side, some of these are obscure to me -- I could not guess, so I asked Gemini which seems pretty good at decoding:
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
Our Op-Ed in @IndianExpress today. Our title for the article was 𝐆𝐨𝐨𝐝𝐡𝐚𝐫𝐝𝐭'𝐬 𝐂𝐮𝐫𝐬𝐞: 𝐈𝐧𝐝𝐢𝐚'𝐬 𝐨𝐛𝐬𝐞𝐬𝐬𝐢𝐨𝐧 𝐰𝐢𝐭𝐡 𝐫𝐚𝐧𝐤𝐢𝐧𝐠𝐬 𝐢𝐬 𝐝𝐚𝐦𝐚𝐠𝐢𝐧𝐠 𝐢𝐭𝐬 𝐚𝐜𝐚𝐝𝐞𝐦𝐢𝐜 𝐞𝐜𝐨𝐬𝐲𝐬𝐭𝐞𝐦.
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
300 papers at 600 USD each. Researchers could have saved 1.8 crore rupees just in this journal if they had checked journaltrends.netlify.app
Note : We made this tool
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
The Curious Case of Max Planck retracted papers. When past scientific practices meet contemporary publishing norms
Yves Gingras, Mahdi Khelfaoui
arxiv.org/abs/2605.17534 [𝚌𝚜.𝙳𝙻 𝚙𝚑𝚢𝚜𝚒𝚌𝚜.𝚑𝚒𝚜𝚝-𝚙𝚑 𝚙𝚑𝚢𝚜𝚒𝚌𝚜.𝚜𝚘𝚌-𝚙𝚑]
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
We have created a tool Journal Trends (link: journaltrends.netlify.app) to help visualize country wise, institute wise and author wise trends for any journal. It can help researchers stay away from predatory journals and choose more wisely.
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
Stealth corrections are still a threat to scientific integrity. If you are not familiar with stealth corrections, please read this.
This guest post is on @deevybee's blog and is authored by from René Aquarius, Floris Schoeters, Alex Glynn and @gcabanac
buff.ly/t8fWFZq
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
"Screening Crossref for integrity issues: Massive and longitudinal mining of publication metadata."
ieeexplore.ieee.org/abstract…
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
Correction to a retraction highlights tortured phrases have been around longer than LLMs.
retractionwatch.com/2026/02/…
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
This now infamous article was only in press for four days before it was retracted. Yet, it is cited more then 80% of other @FrontiersIn articles and has more views than 99% of other #Frontiers articles. It has been viewed almost half a million times.
Not bad for a paper that was only (legitimatly) available for less than a week.
buff.ly/rSRG1K2
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
Is @KimiProduct the new #SCIgen? a paper generator that randomly generates nonsense computer science research papers, based on a single prompt?
en.wikipedia.org/wiki/SCIgen
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
“Thanks for your kind letter and invitation, though”
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
MDPI makes three "stealth corrections" of the peer review record after professor flags the papers as "being affected by a review mill."
pubpeer.com/publications/EA4…
linkedin.com/posts/ma-oviedo…
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
8 (!!) years after the problems were put on Pubpeer this paper finally got retracted! 👇 It's just not ok how long it takes to get rid of bad/fake research! Publishers and @C0PE need to wake up. Too much power with the ones that wrote the papers.
pubpeer.com/publications/749…
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
Reverse Image Forensics Challenge: Try to find a unique area in Figure 9 of this recent Scientific Reports paper: 10.1038/s41598-025-17456-6 [Aug 2025] - Annotated by ImageTwin
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
"Je me suis aperçu que, si le travail bien fait est source de joies puissantes, la paresse savourée en gourmet ne l'est pas moins. Les gens qui ne travaillent pas ne s'ennuient jamais. Les gens qui travaillent s'ennuient quand ils ne travaillent pas."
François Cavanna, Les pensées
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
#MiniReviewMill 58 #FakeReview #CoerciveCitation
Reviewers
🔹Monica Butnariu➕Ionel Samfira➕Sarac Ioan University @life_timisoara 🇷🇴
🔹Anonymous
🔷24 verbatim review reports
🔷47 Monica Butnariu’s works to be cited
H/T Reese Richardson (database)➕another sleuth (key)
🧵
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
#ReviewLikeATweet #StrangerThings @MDPIOpenAccess
Round 1
🥇 4 words @Sensors_MDPI
🥈 10 words @Symmetry_MDPI
🥉23 words @diagnostic_mdpi
📜 37 words @Applsci
Is this compatible with a thorough review of the original manuscript?
⬇️(links and more)
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
Je reviens donc sur la bonne nouvelle: notre papier sur les sneaked references (asistdl.onlinelibrary.wiley.…) avec @gcabanac A. Magazinov, et C. Labbé nous a fait obtenir un best paper award. Notre réaction ici: asist.org/2025/08/13/sneaked…
Merci à @LEXPRESS et @GarciaVictor_ pour l'interview sur mon travail de sleuth célébré dans @Nature (nature.com/articles/d41586-0… & bmcmedresmethodol.biomedcent… & researchintegrityjournal.bio…) et notre récent best paper award sur les sneaked references (asistdl.onlinelibrary.wiley.… & asist.org/programs-services/…).
Vous pouvez lire l'interview ici:
lexpress.fr/sciences-sante/l…
Guillaume Cabanac ⟨here and elsewhere⟩ retweeted
Les tortellinis mis à toutes les sauces : « Guillaume Cabanac @gcabanac lui, développe les outils de détection automatique, par exemple le "Problematic Paper Screener", un logiciel qui repère les
"tortellinis" , ces expressions tordues ou mots inventés qui résultent de tentatives automatisées de masquer un plagiat. »
« Lonni Besançon @lonnibesancon , le détective de la science qui traque les dérives : "Le système pousse à tricher" »
lexpress.fr/sciences-sante/l… @GarciaVictor_
